A data scientist hacked Spotify to determine the saddest hits of all time

Mope up your next playlist with these scientifically-sad songs.

August 24, 2018

Sometimes we all need a good wallow in our own misery, and what better way to kickstart your slide down that sadness-hole than with the perfect mopey soundtrack? Now, thanks to the hacking ways of a British data scientist, your next moody playlist can be scientifically-certified as Sad.

Reporting for the BBC, Miriam Quick took a deep dive into Spotify’s API, which utilizes a number of parameters to categorize the vibes held by various songs – among them a “valence index,” a decimal score ranging from 0 (A Crow Looked At Me levels of heart-crushing sadness) to 1 (“Happy”-er than even Pharrell could muster).

Spotify’s valence parameter has been hacked before to answer such questions as “What’s the most depressing Christmas song?” but Quick’s undertaking was far more ambitious – take every Billboard #1 since 1958 (that’s 1080 tracks, FYI) and match them to their valence score, as determined by Spotify, in order to objectively crown the saddest hits of all time.

Without further ado, the five saddest number one hits since 1958 are…

  1. 5. Commodores – “Still” (1978)

  1. 4. Larry Verne – “Mr. Custer” (1960)

  1. 3. Elvis Presley – “Are You Lonesome Tonight?” (1960)

  1. 2. Commodores – “Three Times a Lady” (1978)

  1. 1. Roberta Flack – “The First Time Ever I Saw Your Face” (1969)

Exclusive videos, interviews, contests & more.

sign up for the a.side newsletter

sign up